Stacking FLATS multiple nights

[Baskevo]

Hi yall,

Quick question: I'm going to be stacking from multiple nights, and I took flats/dark flats on each night. If I run Pixinsight's Weighted batch PP script, can I just throw in all the frames as if I were stacking one night? Or do I have to manually calibrate each night's light frames with that nights flat frames?

Thanks in advance!

[JimMinCT]

Did you change the optical train in any way? If so, then yes, you have to shoot each night and calibrate separately.
If your setup doesn't change from night to night, you can usually get away with a single batch.

[Baskevo]

Did you change the optical train in any way? If so, then yes, you have to shoot each night and calibrate separately.
If your setup doesn't change from night to night, you can usually get away with a single batch.

Well I tried not to change anything, but I noticed some vignetting in the last night, which I didn't notice in the first night... so I think just to be safe I will calibrate them separately.

So can I just run each night and each filter separately through the batch processing script, and select 'calibrate only'? And then run everything together through again with the integration? Or do I have to create all the master darks, flats, and dark flats, calibrate, and integrate myself?

[JimMinCT]

That's what I do. Calibrate the flats, darks and bias frames using the BPP script, and then manually finish using the calibration, registration, and integration tools.

[UlteriorModem]

Yea to be safe they should be calibrated separately. But frankly if you do not change your camera orientation, filters, optical assembly it would probably be fine.

The dust motes will probably rotate with the camera so in that aspect rotation is irrelevant. However if you have any vigenetting that will NOT rotate with the camera. So if the camera is rotated between sessions, even by a few degrees, it will cause minor issues.

The main thing is the camera orientation. I put masking tape on all the rotating parts of my optical train. Mark them with a marker as an index. Then slice the tape where things come apart. This way I can get the camera in the same orientation consistently and dont have to worry about shooting flats every session.

Just create one master flat and use that for a while until something changes.

[Baskevo]

Thanks guys. I'm going to calibrate separately just to be safe My field flattener likes to slip out of the focuser collar a tiny bit, so I'm sure it is rotated a little bit when I have to push it back in even though I put it back in the same spot.

[Baskevo]

I have to run through the calibration 8 times T.T lol

So after I run through the calibration in WBPP script and I have my calibrated light frames, do I run through WBPP again and just not load the flats, darks, and dark flats, and deselect 'calibrate only'? Or should I run it through with dark frames again? Or should I just load up all of the calibration frames with the calibrated light frames? lol this can get complicated

[UlteriorModem]

Not following you. If your batch processing you load the matching lights, darks, bias, and flats and let it do its thing, just don't register them.

Once you have Your calibrated output for each matched set, run star alignment, registration on the lot of them register them to one another.

Once you have the registered output run image integration.

[Baskevo]

So I'm running the calibration only through batch processing, right? It outputs to a calibrated lights folder. Now I need to register and integrate the files. I want to throw the calibrated lights back into weighted batch PP script. When I load the calibrated light frames in the WBPP script (because I want the weights) to register and stack the frames, do I reload the calibration frames (darks, flats, etc) along with the already-calibrated light frames? Or, do I ONLY load the already-calibrated light frames in the WBPP script, without any calibration frames, to register and stack the frames?

If that doesn't make sense, let me explain what I'm doing.
I have about 4 nights of data, 3 different filters.
Each night, I ran a separate set of flats and dark flats for each filter.

so I calibrate only with the Ha light frames, flats, dark flats, and darks. It outputs the calibrated light frames. Now, I want to register and stack WITH the WBPP script, for the functionality of the script. Do I reload the calibrated light frames to register and integrate in the WBPP script WITHOUT the calibration frames, because I've already calibrated and don't need them anymore? Or is there a benefit and/or need to reload the calibration frames with the newly-calibrated Ha light frames?

[UlteriorModem]

Just have the BPP calibrate the matching frames or run the calibration process and skip the script, then use star alignment process on them (not the script), once registered stack them.

I know the script is handy but it seems like it is complicating things for you with all the different data sets.

[JimMinCT]

The key difference here is not to use the script the final time.
Register and integrate using the respective tools, not the script.

[Baskevo]

Hmm... so I would have to do the weighting myself with the subframe selector process?

I don't know, I feel like there is a way to do it... If I don't add the calibration frames on the second time through and unselect "calibrate only" so it registers and integrates WITH the weights (which is my whole reason for using the script), then shouldn't it just skip the calibration? I'm going to try it... I'll let you know how it goes

[Stuart]

Try it both ways and see what you get.

[Baskevo]

Try it both ways and see what you get.

Thats a good plan. I have 4 different things to try... I have 2 sets of flats for 3 different filters, 2 different processes... It's gonna be a long night lol

[twideman]

Late to the thread, but found while having the same issue. Here's what I'm doing:

I have a six-panel mosaic, shot on seven nights, L,R,G,B+Ha, even different exposure lengths. Seven sets of flats. Some of the shots (for instance, Lum on panel 3) spanned multiple nights.

I, too, want to take advantage of the WBPP features, especially weighting.

I processed the images using WBPP by each night (i.e. by each flat set), allowing auto selection of registration, everything. I ran WBPP separately for each exposure length as I was using CosmeticCorrection with a dark frame of matching exposure length. I saved the output results to a text file for each mosaic frame, appending more text each time it was generated. This resulted in multiple registration reference frames due to the multiple passes, and registered frames with WBPPWEIGHT data in the FITS headers. Note that these "registered" frames are not registered all together, but to the auto-selected frame for the set for each WBPP pass.

I then went back to the text file and noted which files were chosen as the reference frames from the various WBPP passes. I opened these registered reference frames (the _r version) in PI and looked at the FITS headers for WBPPWEIGHT. For each mosaic frame, I chose the one reference frame with the highest WBPPWEIGHT as I figured it was a safe bet for the entire mosaic frame reference.

I then ran StarAlignment, selecting this one frame as the reference for all exposures for that particular mosaic frame, and selecting the _r (registered) frames from the WBPP output as the input files for alignment. I output to a new "reregistered" folder with suffix _rr, and used these reregistered frames -- which contained WBPPWEIGHT from their initial WBPP pass -- in ImageIntegration to generate the master frames.

So far, works for me... Best of both worlds. Specific flats by night, the ability to span nights, and the chance to use WBPPWEIGHT in ImageIntegration.

I'd love to hear why I'm doing it wrong, as I just kinda sussed this out as a possible solution and would love to hear feedback.